mihirp1998 / AlignProp
AlignProp uses direct reward backpropogation for the alignment of large-scale text-to-image diffusion models. Our method is 25x more sample and compute efficient than reinforcement learning methods (PPO) for finetuning Stable Diffusion
☆273Updated 4 months ago
Alternatives and similar repositories for AlignProp:
Users that are interested in AlignProp are comparing it to the libraries listed below
- Code for "Diffusion Model Alignment Using Direct Preference Optimization"☆364Updated 3 weeks ago
- [CVPR 2024] Code for the paper "Using Human Feedback to Fine-tune Diffusion Models without Any Reward Model"☆196Updated 10 months ago
- [ICCV 2023] Efficient Diffusion Training via Min-SNR Weighting Strategy☆240Updated 2 months ago
- Smooth Diffusion: Crafting Smooth Latent Spaces in Diffusion Models arXiv 2023 / CVPR 2024☆331Updated 5 months ago
- [Neurips 2023 & TPAMI] T2I-CompBench (++) for Compositional Text-to-image Generation Evaluation☆236Updated last month
- GenEval: An object-focused framework for evaluating text-to-image alignment☆181Updated 7 months ago
- 🔥 [CVPR2024] Official implementation of "Self-correcting LLM-controlled Diffusion Models (SLD)☆166Updated 10 months ago
- (CVPR 2024) 🧩 TokenCompose: Text-to-Image Diffusion with Token-level Supervision☆120Updated 2 months ago
- [NeurIPS 2024] ReNO: Enhancing One-step Text-to-Image Models through Reward-based Noise Optimization☆124Updated last month
- Official implementation of "Controlling Text-to-Image Diffusion by Orthogonal Finetuning".☆286Updated 4 months ago
- MoVQGAN - model for the image encoding and reconstruction☆219Updated last year
- [WACV 2024] Training-Free Layout Control with Cross-Attention Guidance☆250Updated 11 months ago
- [NeurIPS 2024] CV-VAE: A Compatible Video VAE for Latent Generative Video Models☆266Updated 2 months ago
- Video Diffusion Alignment via Reward Gradients. We improve a variety of video diffusion models such as VideoCrafter, OpenSora, ModelScope…☆235Updated 6 months ago
- Aesthetic Post-Training Diffusion Models from Generic Preferences with Step-by-step Preference Optimization☆181Updated 2 months ago
- ☆480Updated 2 months ago
- An in-context conditioning version of MUSE with pre-trained checkpoints.☆111Updated last year
- Better Aligning Text-to-Image Models with Human Preference. ICCV 2023☆274Updated last year
- [ICLR 2025] Rectified Diffusion: Straightness Is Not Your Need☆188Updated 2 months ago
- Code for the paper "Training Diffusion Models with Reinforcement Learning"☆401Updated last year
- SpeeD: A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training☆176Updated last month
- Official pytorch implementation of the paper: "An Edit Friendly DDPM Noise Space: Inversion and Manipulations". CVPR 2024.☆317Updated 7 months ago
- 🚀 Cross attention map tools for huggingface/diffusers☆226Updated last month
- ☆101Updated last month
- Official PyTorch implementation of the paper "In-Context Learning Unlocked for Diffusion Models"☆385Updated 11 months ago
- [ICLR 2025] OpenVid-1M: A Large-Scale High-Quality Dataset for Text-to-video Generation☆247Updated this week
- Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesis☆455Updated 9 months ago
- Implementation of "DiffFit: Unlocking Transferability of Large Diffusion Models via Simple Parameter-Efficient Fine-Tuning"☆82Updated last year
- [ICCV 2023] Official PyTorch implementation for the paper "FreeDoM: Training-Free Energy-Guided Conditional Diffusion Model"☆281Updated last year
- VisionReward: Fine-Grained Multi-Dimensional Human Preference Learning for Image and Video Generation☆153Updated 2 weeks ago